A Logical Framework for Graph Theoretical Decision Tree Learning

Abstract. We present a logical approach to graph theoretical learning that is based on using alphabetic substitutions for modelling graph morphisms. A classi ed graph is represented by a de nite clause that possesses variables of the sort node for representing nodes and atoms for representing the edges. In contrast to the standard logical semantics, different node variables are assumed to denote di erent objects. The use of an alphabetical subsumption relation ( -subsumption) implies that the least generalization of clauses ( -generalization) has di erent properties than Plotkin's least generalization (lgg). We present a method for constructing optimal -generalizations from Plotkin's least generalization. The developed framework is used in the relational decision tree algorithm TRITOP.

• Type: Conference
• Year: 1997
• Where: ILP
• Authors: Peter Geibel, Fritz Wysotzki
• Keywords: Automated Reasoning, Classi Ed Graph, Di Erent Objects, ILP 1997, Standard Logical Semantics